Apple's AI Delay, Meta's Llama 3.1, Google Gemini's Upgrade, Elon Musk's Grok AI Ethical Dilemmas

📰 Welcome to this week's edition of our AI and Tech Newsletter!

Table of Contents

1. 📱 Apple Delays AI Features for iOS 18.1

2. 🤖 Meta's Llama 3.1: Pioneering Open Source AI

3. 🌍 Google Gemini's Big Upgrade

4. 📹 Runway AI Video Generator's Training Controversy

5. 🛠️ Anthropic's Web Crawler Scrutiny

6. 🔍 OpenAI Tests SearchGPT Prototype

7. 🏛️ Apple Commits to White House AI Safety Pledge

8. 🔐 Privacy Concerns with Musk's Grok AI Training

9. ❌ How to Disable X's (Twitter) Data Sharing

10. 🔍 OpenAI Unveils SearchGPT

11. 💼 ZoomInfo Alum Raises $15M for AI Sales Engineers

12. 🌟 Mistral's Large 2 Model Rivals OpenAI and Meta

13. 📈Trending Repos

1. 📱 Apple Delays AI Features for iOS 18.1

Apple Intelligence AI Features to Wait for iOS 18.1

Apple's much-anticipated AI features, including a "smarter Siri" and ChatGPT integrations, will not launch with iOS 18 and the iPhone 16 in September. Instead, these features will arrive with iOS 18.1 in October to ensure a smooth rollout. This delay allows developers to test the new AI capabilities and iron out any bugs before the public release.

Key Highlights:

• Delayed Features: Siri improvements, AI-powered image generation, and ChatGPT integration.

• Developer Beta: Early access to iOS 18.1 for testing and feedback.

• Future Updates: Full functionality expected by spring 2025.

🔗 Read More

2. 🤖 Meta's Llama 3.1: Pioneering Open Source AI

Meta Advances Open Source AI with ‘Frontier-Level’ Llama 3.1

Meta has unveiled Llama 3.1, marking a significant milestone in open-source AI. This release includes three models: 405B, 70B, and 8B, with the 405B model competing with the most advanced closed models while offering better cost-efficiency.

Key Highlights:

• Customization: Organizations can train and fine-tune models with specific data.

• Cost-Efficiency: Llama 3.1 405B runs at roughly half the cost of closed models like GPT-4.

• Partnerships: Collaboration with major tech companies for development services.

🔗 Read More

3. 🌍 Google Gemini's Big Upgrade

Gemini’s Big Upgrade: Faster Responses with 1.5 Flash, Expanded Access and More

Google has upgraded its Gemini AI to version 1.5 Flash, offering faster and more helpful responses across more than 40 languages and 230 countries. The update also includes a new related content feature to combat hallucinations and expanded access for teenagers globally.

Key Highlights:

• Faster Responses: Enhanced reasoning and image understanding.

• Expanded Access: Available to teens and in more languages.

• Related Content Links: Helps reduce hallucinations and provides additional information sources.

🔗 Read More

4. 📹 Runway AI Video Generator's Training Controversy

AI Video Generator Runway Trained on Thousands of YouTube Videos Without Permission

A leaked document reveals that Runway's AI video generation tool was trained using thousands of YouTube videos and pirated films without permission. This has sparked significant controversy and raised ethical concerns about data usage in AI training.

Key Highlights:

• Unethical Training: Use of unlicensed YouTube videos and pirated films.

• Industry Reaction: Questions about the transparency and ethics of AI training data.

• Company Response: Runway has yet to provide specifics about its data sources.

🔗 Read More

5. 🛠️ Anthropic's Web Crawler Scrutiny

Anthropic’s Crawler Ignores Websites’ Anti-AI Scraping Policies

Anthropic's web crawler, used for training its AI models, has been found violating websites' terms of use by scraping data without permission. iFixit and other websites reported excessive traffic and unauthorized data usage, sparking debates on AI data collection ethics.

Key Highlights:

• Policy Violations: Scraping data despite explicit prohibitions.

• Impact: High traffic causing server issues for affected websites.

• Company Response: Adjusted scraping policies following complaints.

🔗 Read More

6. 🔍 OpenAI Tests SearchGPT Prototype

OpenAI Tests SearchGPT, a New AI-Powered Search Prototype

OpenAI is testing SearchGPT, a prototype search feature that combines AI models with web information to provide timely and accurate answers. The prototype aims to enhance search efficiency and accuracy while maintaining clear and relevant source attribution.

Key Highlights:

• Conversational Search: Real-time information from the web with AI enhancements.

• Publisher Partnerships: Collaborating with publishers to manage content appearance.

• User Feedback: Gathering insights to improve the search experience.

🔗 Read More

7. 🏛️ Apple Commits to White House AI Safety Pledge

Apple Signs the White House’s Commitment to AI Safety

Apple has signed the White House's voluntary commitment to developing safe, secure, and trustworthy AI. This move signals Apple's dedication to ethical AI development as it integrates generative AI into its core products, including the upcoming Apple Intelligence.

Key Highlights:

• AI Safety Pledge: Commitment to developing secure and trustworthy AI.

• Generative AI Integration: Plans to embed AI across Apple's product ecosystem.

• Regulatory Alignment: Aligning with federal guidelines to preempt regulatory challenges.

🔗 Read More

8. 🔐 Privacy Concerns with Musk's Grok AI Training

Privacy Watchdog 'Surprised' by Musk Opting User Data into Grok AI Training

X, formerly Twitter, has defaulted user data into its Grok AI training pool without explicit user consent. This move has drawn criticism from privacy watchdogs and raised concerns about compliance with data protection regulations.

Key Highlights:

• User Data Usage: Default opt-in for AI training without clear user consent.

• Regulatory Scrutiny: European privacy watchdogs investigating compliance issues.

• User Actions: How to opt-out and protect personal data.

🔗 Read More

9. ❌ How to Disable X's (Twitter) Data Sharing

Here’s How to Disable X (Twitter) from Using Your Data to Train Its Grok AI

X has activated a setting that allows the company to train its Grok AI using users’ posts by default. Users can disable this setting and delete their conversation history with the AI to protect their data.

Key Highlights:

• Data Sharing: Default-enabled setting for AI training.

• User Control: Steps to disable data sharing and delete conversation history.

• Privacy Protection: Ensuring user data security.

🔗 Read More

10. 🔍 OpenAI Unveils SearchGPT

With Google in Its Sights, OpenAI Unveils SearchGPT

OpenAI has launched SearchGPT, an AI-powered search feature designed to deliver timely and accurate answers with clear source attribution. This prototype aims to challenge traditional search engines by integrating real-time web information with conversational AI capabilities.

Key Highlights:

• AI-Powered Search: Combining AI models with real-time web data.

• User Experience: Enhanced search efficiency with clear source links.

• Future Integration: Plans to incorporate SearchGPT features into ChatGPT.

🔗 Read More

11. 💼 ZoomInfo Alum Raises $15M for AI Sales Engineers

ZoomInfo Alum Raises $15M for Startup That Builds AI Sales Engineers

Arjun Pillai, a former ZoomInfo executive, has raised $15 million for his new startup, DocketAI. The company develops virtual sales engineers to assist non-technical salespeople with technical queries and documentation, improving sales efficiency.

Key Highlights:

• AI Sales Engineers: Virtual assistants for technical sales support.

• Funding Success: $15 million Series A led by Mayfield and Foundation Capital.

• Market Impact: Enhancing productivity and win rates for sales organizations.

🔗 Read More

12. 🌟 Mistral's Large 2 Model Rivals OpenAI and Meta

Mistral’s Large 2 is its Answer to Meta and OpenAI’s Latest Models

Mistral has released its new flagship model, Large 2, which rivals the latest models from OpenAI and Meta. With advanced capabilities in code generation, mathematics, and reasoning, Large 2 aims to set new performance and cost standards for open AI models.

Key Highlights:

• Advanced Capabilities: Outperforms Llama 3.1 in several benchmarks.

• Cost-Efficiency: High performance with fewer parameters.

• Widespread Access: Available on major cloud platforms and Mistral’s own platform.

🔗 Read More

13. 📈Trending Repos

📚 Bark

GitHub Repository: Bark

Description: Bark is an advanced text-to-speech (TTS) model by Suno AI, designed to generate natural-sounding speech from text. With its high-quality voice synthesis and multi-language support, Bark is ideal for adding realistic speech features to your applications. 🗣️🔊

Highlights:

• High-Quality Speech Generation: Produces clear, natural-sounding speech🎙️.

• Multi-Language Support: Supports various languages and accents🌍.

• Open-Source: Customizable and integrable into different projects🛠️.

🌍 CodeCarbon

GitHub Repository: CodeCarbon

Description: CodeCarbon helps measure and reduce the carbon footprint of machine learning models. It provides valuable metrics on emissions associated with ML training, enabling developers to make eco-friendly choices. 🌱💻

Highlights:

• Carbon Emission Metrics: Offers insights into the carbon footprint of ML processes📉.

• Framework Integration: Compatible with popular ML frameworks like TensorFlow and PyTorch🔗.

• Environmental Impact Awareness: Promotes sustainability in machine learning🌿.

Stay tuned for more updates and breakthroughs in the AI and tech world! 🚀💡

Follow us on social media and never miss an update!

📧 Have questions or feedback? Reply to this email—we’d love to hear from you!

🔗 Unsubscribe from this newsletter if you no longer wish to receive it.